The summer program we propose here has as its objective to adjust and test our prediction algorithm for use in real life situations. We will first run the code for a realistic class imbalance…

This is a $54,240.00 grant for research awarded by the Shared Services Center to the Georgia State University Research Foundation Inc., a non-profit organization and foundation supporting Georgia State University in Atlanta, Georgia. The grant will fund a summer program to adjust and test a prediction algorithm for use in real-life situations, with the goal of integrating the operational code into the CCMC and SRAG systems for which the Georgia State University Research Foundation Inc. submitted a proposal in November 2023. The grant has an ultimate completion date of March 28, 2025 and does not have a set-aside designation.

The Georgia State University Research Foundation Inc. has a history of providing research, education, training, and program implementation services through both prime and sub federal contracts and grants across areas like science, technology, engineering, healthcare, education, criminal justice, and social services. This includes past prime contracts such as a $6.3 million firm fixed price delivery order contract from the Department of Health and Human Services Health Resources and Services Administration (HRSA) and a $2.99 million project grant from the Department of the Air Force Materiel Command Research Laboratory.

The summer program we propose here has as its objective to adjust and test our prediction algorithm for use in real life situations. We will first run the code for a realistic class imbalance (non-events/events > 10:1, instead of 1:1) and in a simulated real-time setting, i.e. train the classifier for only events in the past, and then apply IT to the near future, just as in operational forecasting. Once IT has been demonstrated that the code is very accurate in these circumstances, we can prepare our algorithm for use integration into the ccmc and srag systems, for which we have submitted a year 3 proposal in november 2023. Our operational code will have very low computing requirements - IT can easily keep up on a standard macbook air. Hence its operation by the ccmc and srag would require few resources. We intend to maintain copies of the forecasting systems and gsep at GSU to achieve further optimization, keep gsep current, and be ready to assist our NASA partners in case of software issues.